Sign up for our newsletter and get the latest HPC news and analysis.
Send me information from insideHPC:


Job of the Week: PhD scholarship in GPU processing for Super Resolution Ultrasound Imaging

The Technical University of Denmark (DTU) is seeking a in our PhD scholarship in GPU processing for Super Resolution Ultrasound Imaging in our Job of the Week.

A 3-year PhD scholarship is available at the Center for Fast Ultrasound Imaging (CFU), Department of Health Technology from March 2020 sponsored by the European Research Council (ERC) and the Department of Health Technology at DTU. It is conducted in collaboration between CFU and the MEMS group at DTU Health Tech, the University of Copenhagen, and the Copenhagen University Hospital (Rigshospitalet) in the ERC Synergy Grant “3-D Super resolution Ultrasound Real time imaging of Erythrocytes” – SURE.

The SURE project develops and researches a new super resolution ultrasound imaging method capable of resolving 3-D capillary flow in the human body. The approach tracks the motion of the individual red blood cells (erythrocytes) in a three-dimensional volume for a full visualization of anatomy, flow, and perfusion. The approach can potentially yield a paradigm shift in the scientific study, diagnoses, and treatment of cancer, diabetes, and vascular diseases at the capillary level, as it enables the possibility of volumetrically visualizing capillary perfusion in real-time at frame rates above 20 Hz without injection of contrast agents.

Responsibilities and tasks

This PhD projects develops real time processing of the acquired data using GPUs. The research is conducted in a truly multi-disciplinary environment consisting of the four academic partners specializing in advanced ultrasound imaging, silicon CMUT probes, animal experiments, and clinical investigations, and the main purpose is to translate the SURE method from a research idea to the clinic.

The main task is to translate the Matlab based SURE processing framework to a real-time implementation using GPUs. Data from the scanners are acquired at several Gbytes per second, and this should be beamformed and processed by a GPU cluster in real time. The overall goal is to speed up research and finally present the SURE volume in real time. The processing pipeline will be based on our current matrix GPU beamformer, which will be optimized and integrated with a GPU implementation of the SURE processing framework. 3-D visualization will also be conducted by the GPUs to make it easy to navigate the 3-D volumes, which often contain more than 100 Gbytes of data. The full scheme will be implemented on our research scanners and on a Nvidia GPU cluster for use in human trials conducted together with the Department of Radiology at Rigshospitalet, where real-time performance and usability will be highly important.

Looking for a new gig? Our Jobs Board helps companies of all sizes hire the best talent and offers the best opportunity for job seekers to get hired.

Are you paying too much for your job ads? Priced at just $99.99 dollars for 90 days, ads on our insideHPC Jobs board are a great way to reach the top supercomputing professionals.

Post a Job 

Resource Links: